CJC - Cleveland

Below are select errors from match against CJC from Cleveland earlier this year.


***************

I trail 8 away 7 away and have 41 to play.

is CJC score: 2 pip: 111

9 point match

pip: 135 score: 1 is Bill

XGID=-a--aCDAB---bD-----cbAbbb-:0:0:1:41:1:2:0:9:10

to play 41

1.	XG Roller++	8/4* 5/4	eq: -0.093
	Player: Opponent: 47.10% (G:14.58% B:0.84%) 52.90% (G:34.38% B:0.91%)
2.	XG Roller++	21/17 7/6	eq: -0.757 (-0.664)
	Player: Opponent: 34.73% (G:5.42% B:0.27%) 65.27% (G:27.21% B:0.15%)
3.	2-ply	8/7 8/4*	eq: -0.430 (-0.337)
	Player: Opponent: 41.37% (G:9.80% B:0.47%) 58.63% (G:34.66% B:0.41%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I made the monster blunder 21/17, 8/7. Once again I was paralyzed by fear and wanted to keep the blot count down.

The best play was 8/4*, 5/4. Clearly this has the most upside. Blots be damned, this plays to win rather than playing not to lose. Over and over again lately I am seeing this theme in my games. Note how the wins get jacked up 12% by making the 4 point on her head. On the other hand, the gammons are pretty bad either way. The big play gets gammoned 34% while the wimpy play still gets gammoned 27% anyway.

When in doubt make the DMP play I suppose.

**************

I lead 6 away 7 away and have 42 to play.

is CJC score: 2 pip: 138

9 point match

pip: 154 score: 3 is Bill

XGID=---aBBC-C---bB-bbb-f--B-A-:0:0:1:42:3:2:0:9:10

to play 42

1.	XG Roller++	24/20 22/20	eq: +0.123
	Player: Opponent: 50.46% (G:15.36% B:0.42%) 49.54% (G:7.25% B:0.26%)
2.	XG Roller++	24/18	eq: 0.000 (-0.124)
	Player: Opponent: 48.17% (G:14.57% B:0.45%) 51.83% (G:9.78% B:0.37%)
3.	3-ply	22/18 8/6	eq: -0.026 (-0.149)
	Player: Opponent: 48.67% (G:15.74% B:0.62%) 51.33% (G:12.82% B:0.62%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I moved 24/18. The idea has some merit - trying to force contact when she is outboarded. But this is the wrong concept.

Advancing the anchor 24/20, 22/20 is the right idea. I establish the best of the "anchor and gaurd" setups. I am down in the race so the man on 24 point must stay. Also by inching up the 5 point my anchor is not impeded (or blocked six away) which is a nice tip I learned from Julius High recently. Finally by making the 5 anchor I devalue the six stacked men on CJC's six point. These checkers can not now be put to use to make points in front of my anchor.


**************

Score is 6 away 6 away and I am on roll. Cube action?

is CJC score: 3 pip: 125

9 point match

pip: 126 score: 3 is Bill

XGID=---aaaEBCB--bBA----b-cbc--:0:0:1:00:3:3:0:9:10

on roll, cube action?

Analyzed in Rollout	No double	Double/Take
Player Winning Chances:	71.35% (G:12.24% B:0.56%)	71.43% (G:12.09% B:0.62%)
Opponent Winning Chances:	28.65% (G:4.06% B:0.00%)	28.57% (G:4.12% B:0.00%)
Cubeless Equities	+0.527	+1.130
Cubeful Equities
No double:	+0.798 (-0.115)	±0.019 (+0.779..+0.817)
Double/Take:	+0.913	±0.026 (+0.888..+0.939)
Double/Pass:	+1.000 (+0.087)
Best Cube action: Double / Take
Rollout details
385 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller

I rolled.  It's been so long I can't remember my thought process.  Probably I noticed the race was close, I am outboarded, have a blot to cleanup, and I have only a one-point board....let's take a roll.

True enough.  But let's look at the position another way.  Carol's board is overrun with an open 5-point.  Should an accident happen, Carol won't have anything until the 5 point is filled in.

I am up in the race (not much but up) and have an outside prime.  So I may just win a simple race.  I have significant blitz and gammon potential here.  She has three vulnerable blots which could get scooped up if things go well.  From this perspective, the double is clear and the take a tough one.

****************

On the other hand, if you double then you have to know how to play the rolls.  How to handle this 21?

is CJC score: 3 pip: 125

9 point match

pip: 126 score: 3 is Bill

XGID=---aaaEBCB--bBA----b-cbc--:0:0:1:21:3:3:0:9:10

to play 21

1.	XG Roller++	6/5* 5/3*	eq: +0.544
	Player: Opponent: 65.28% (G:12.97% B:0.69%) 34.72% (G:10.84% B:0.23%)
2.	XG Roller++	14/13 8/6	eq: +0.423 (-0.122)
	Player: Opponent: 62.06% (G:6.50% B:0.24%) 37.94% (G:4.29% B:0.06%)
3.	3-ply	6/4* 4/3*	eq: +0.447 (-0.097)
	Player: Opponent: 63.24% (G:12.34% B:0.55%) 36.76% (G:11.54% B:0.20%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I played the hyper-wimpy 14/13, 8/6.  Idea was to ease up on a cube.  The problem is I am giving Carol the whole roll to do something good.  Little numbers anchor up while big numbers run.

The double-hit 6/5*/3* gives Carol less good rolls.  It slows her down in the race and gives here less anchoring rolls.  Also the hits aren't too devastating either.  As noted above, she has the open 5-point and will have three men behind a 4-prime even if she hits.

***************

I trail 6 away 2 away and have 33 to play.

is CJC score: 7 pip: 121

9 point match

pip: 67 score: 3 is Bill

XGID=aCCBB-C---Aab---Ab-cbb-b--:1:-1:1:33:3:7:0:9:10

to play 33

1.	XG Roller++	16/4	eq: +1.226
	Player: Opponent: 92.49% (G:36.78% B:0.04%) 7.51% (G:0.16% B:0.01%)
2.	XG Roller++	16/13 10/1	eq: +1.045 (-0.180)
	Player: Opponent: 83.58% (G:36.43% B:0.09%) 16.42% (G:0.73% B:0.03%)
3.	XG Roller+	16/7 10/7	eq: +1.199 (-0.026)
	Player: Opponent: 92.17% (G:34.89% B:0.02%) 7.83% (G:0.12% B:0.00%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I tried to force the gammon, moving 16/13, 10/1. This trades three extra jokers (51, 52 or 3 more hits than the safe play) for the chance to pick up a second blot.

Turns out the relatively safe 16/4 was best. This gives only 55 as a joker. As you can see by looking at the numbers I win the same amount of gammons either way so less shots is the most important thing. CJC has 8 extra outside crossovers besides the man on the bar. Hard to estimate the gammons over the board, but suffice to say it's a lot.

Two Trice bearoff cubes from the chouette

At the chouette this week we had both (a) Stack & Straggler and (b) Rolls versus Pips cube actions. I find these happen every so often, but just infrequently enough that my memory is hazy as to all the details.

Take a look at my earlier post for reference on EPC basics.

http://bigmoneybill.blogspot.com/2011/12/racingepc-and-prop-57-practice.html

*************************

Rolls versus Pips. Cube Action?

is Player 2 score: 0 pip: 23

Unlimited Game Jacoby Beaver

pip: 30 score: 0 is Player 1

XGID=--AB-BB--------------cbac-:1:1:1:00:0:0:3:0:10

on roll, cube action?

Analyzed in XG Roller++	No redouble	Redouble/Take
Player Winning Chances:	69.45% (G:0.00% B:0.00%)	69.44% (G:0.00% B:0.00%)
Opponent Winning Chances:	30.55% (G:0.00% B:0.00%)	30.56% (G:0.00% B:0.00%)
Cubeless Equities	+0.389	+0.778
Cubeful Equities
No redouble:	+0.661
Redouble/Take:	+0.651 (-0.010)
Redouble/Pass:	+1.000 (+0.339)
Best Cube action: No redouble / Take
Percentage of wrong pass needed to make the double decision right: 2.8%

eXtreme Gammon Version: 2.10

You can see it is borderline No Redouble / Take. Let's try the Trice formula.

EPC: (Pips) = 30 pips + about 7 for wastage or 37

Point of Last Take = (# Rolls to go) - 3
= approx (37 pips / 7) - 3
= about 2 pips

So point of last take = 39 pips for this position

EPC (Rolls) = 7N+1 plus adjustment for nearmiss

= (7*5 + 1) + near miss adjustment
= 36 + near miss adjustment

near-miss: Avg Pip is 23 pips / 9 checkers or 2.5
2.5 translates to 1 pip adjustment.

So EPC (rolls) = 36 + 1 = 37 pips

Application:

Double w/in 2 pips of last take
Redouble w/in 1 pip of last take

Implies that this position would be an initial double but not a redouble.

Appears to be right on target! Pretty neat.


*******************

Stack & Straggler vs. Stack & Straggler:  Cube Action?

is Player 2 score: 0 pip: 20

Unlimited Game Jacoby Beaver

pip: 32 score: 0 is Player 1

XGID=-CBBB------A-a----------h-:2:-1:-1:00:0:0:3:0:10

on roll, cube action?

Analyzed in Rollout	No redouble	Redouble/Take
Player Winning Chances:	77.69% (G:0.00% B:0.00%)	77.68% (G:0.00% B:0.00%)
Opponent Winning Chances:	22.31% (G:0.00% B:0.00%)	22.32% (G:0.00% B:0.00%)
Cubeless Equities	+0.554	+1.107
Cubeful Equities
No redouble:	+0.813 (-0.161)	±0.002 (+0.811..+0.816)
Redouble/Take:	+0.974	±0.005 (+0.969..+0.979)
Redouble/Pass:	+1.000 (+0.026)
Best Cube action: Redouble / Take
Rollout details
1296 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller
Double Decision confidence:		100.0%
Take Decision confidence:		100.0%
Duration: 1.1 seconds

eXtreme Gammon Version: 2.10

Computer shows it as Redouble / Take. Let's try Trice's Stack & Straggler method and see how it does.

Recall for Stack + 1 straggler: EPC = (3.5 x # of checkers) + the outside man


EPC (doubler) = (3.5 x 9) + 12
                        = 43.5

Point of Last take =   (43.5 / 7) or about 6 rolls to go less 3
    =  3 pips  or 46.5 pips for taker is point of last take

EPC (taker) =(3.5 x 10) + 11
                    = 46


Application:
Taker is just barely within the point-of last take implies this is a Redouble / Close Take

Once again the method was correct.  Very impressive!

DT

Below are errors from a match against David Todd from the Cleveland tournament earlier this year.

**************

First game of 11 point match and I have 31 to play.

is David Todd score: 0 pip: 133

11 point match

pip: 134 score: 0 is Bill

XGID=--Bb-BE-B---cB----bbbbb-B-:1:1:1:31:0:0:0:11:10

to play 31

1.	XG Roller++	24/23 8/5	eq: -0.861
	Player: Opponent: 19.31% (G:4.01% B:0.13%) 80.69% (G:32.97% B:1.82%)
2.	XG Roller++	6/2	eq: -0.941 (-0.079)
	Player: Opponent: 13.69% (G:2.30% B:0.06%) 86.31% (G:26.26% B:1.55%)
3.	3-ply	24/23 13/10	eq: -0.917 (-0.056)
	Player: Opponent: 18.22% (G:3.52% B:0.15%) 81.78% (G:34.68% B:2.74%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I am totally in give up or gammon save mode here. 6/2 without even thinking about it.


The computer has brass balls and prefers 24/23, 8/5. This makes a move for an upgraded anchor while DT lacks the ammo to close me out and also duplicates the 5's.

Note the gammons lost (6%) are offset by the increased wins (6%). I have seen this theme before but tread carefully. To my mind, this is one of those exceptions that are hard to find. My guess is this maneuver is usually wrong and horribly so when it is wrong.

***************

I trail 11 away 7 away and have 61 to play.

is David Todd score: 4 pip: 144

11 point match

pip: 142 score: 0 is Bill

XGID=-a-AC-D-B--abC-b-bbe----B-:0:0:1:61:0:4:0:11:10

to play 61

1.	XG Roller++	13/7 4/3	eq: -0.055
	Player: Opponent: 43.61% (G:16.05% B:0.40%) 56.39% (G:12.59% B:0.78%)
2.	XG Roller++	8/2 3/2	eq: -0.186 (-0.131)
	Player: Opponent: 40.51% (G:13.87% B:0.27%) 59.49% (G:13.05% B:0.80%)
3.	3-ply	8/2 4/3	eq: -0.141 (-0.086)
	Player: Opponent: 42.11% (G:14.59% B:0.45%) 57.89% (G:14.24% B:1.04%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I made the deuce point with 8/2, 3/2.

Better was the aggressive 13/7, 4/3. I think the best way to find this play is Neil Kaz's maxim "where do you want to fight?"

The 6/4/3 board is a much better combo than grandma's gapped teeth 6/4/2. This places pressure on DT to roll a six immediately. And even if he does, DT only has a one point board so there is plenty of play left in the position.

**************

Later in the same game and I have caught DT on the bar against a 5 point board. Cube action?

is David Todd score: 4 pip: 139

11 point match

pip: 94 score: 0 is Bill

XGID=aBCBB-D-----b--b-cbe----B-:0:0:1:00:0:4:0:11:10

on roll, cube action?

Analyzed in XG Roller++	No double	Double/Take
Player Winning Chances:	60.87% (G:24.88% B:0.06%)	60.26% (G:25.23% B:0.07%)
Opponent Winning Chances:	39.13% (G:5.65% B:0.19%)	39.74% (G:5.76% B:0.17%)
Cubeless Equities	+0.403	+0.849
Cubeful Equities
No double:	+0.582 (-0.097)
Double/Take:	+0.679
Double/Pass:	+1.000 (+0.321)
Best Cube action: Double / Take

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I paused but rolled. I just couldn't believe I was really threatening anything with two men stuck on the ace point and zero connectivity around the board.

Yet this is a significant mistake not to double. I put the position in for money and it is a borderline No double. So this double appears to be mostly score based. Gammons are slight elevated and his recube potential is slightly crippled. Looks like a snore of a take though, especially for DT who never drops anything.

**************

I trail 11 away 7 away. 32 to play. Stiff or builder action?

is David Todd score: 4 pip: 149

11 point match

pip: 155 score: 1 is Bill

XGID=-b--B-D-C---eD-----bbbbAA-:0:0:1:32:1:4:0:11:10

to play 32

1.	XG Roller++	13/8	eq: -0.398
	Player: Opponent: 38.96% (G:10.63% B:0.49%) 61.04% (G:22.50% B:0.90%)
2.	XG Roller++	13/11 13/10	eq: -0.491 (-0.093)
	Player: Opponent: 37.14% (G:10.70% B:0.54%) 62.86% (G:26.02% B:1.83%)
3.	3-ply	13/10 6/4	eq: -0.446 (-0.048)
	Player: Opponent: 37.78% (G:10.52% B:0.55%) 62.22% (G:24.23% B:1.26%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I went for the builders with 13/11,13/10. My thinking was I should take my chances now while David is stiff himself. He has a very strong position, for sure, but stacked on the midpoint and stuck on the ace he can't really hurt me too much unless he hits a shot now. So I figure I should position myself best for counterplay before it is too late.

The computer thinks prudence is called for with 13/8. I find my arguments above pretty convincing so I am not really sure what is going on here. Perhaps I should just sit back and let him come to me. I am turning 9's and 10's into good rolls which would be kind of blah before? Maybe - not sure about that one.

************

Later in same game. Another 32 to play.

is David Todd score: 4 pip: 146

11 point match

pip: 162 score: 1 is Bill

XGID=--aaB-D-C-A-eB-----bbbbBA-:0:0:1:32:1:4:0:11:10

to play 32

1.	XG Roller++	10/8 6/3*	eq: -0.711
	Player: Opponent: 32.50% (G:8.71% B:0.36%) 67.50% (G:28.76% B:2.32%)
2.	XG Roller++	13/11 13/10	eq: -0.804 (-0.093)
	Player: Opponent: 30.31% (G:7.84% B:0.33%) 69.69% (G:26.69% B:1.94%)
3.	3-ply	8/3*	eq: -0.757 (-0.045)
	Player: Opponent: 32.41% (G:8.99% B:0.36%) 67.59% (G:31.48% B:3.03%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

OK so this game makes me queasy after either play but you still need to get the most out of your bad positions too.

I went with the wimpy 13/10, 13/11. I think the problem with this play is I put absolutely no pressure on David whatsoever. He has complete freedom to play all his rolls and they all play well.

Since this is a lopsided score I will be taking deep so I should just buckle up and make the best play to try and win.
Hence the computer's suggestion 10/8, 6/3* This holds the midpoint, unstacks the six point and puts David on the bar. I would retain the best balance of offense and defense while staying in the game for the long-term with this play.

*************

Still later in the same game. A couple of jokers and I am on the verge of a turnaournd. How best ot move this 63?

is David Todd score: 4 pip: 150

11 point match

pip: 137 score: 1 is Bill

XGID=a-AbBCB-B-Bab-----AbcbbB--:1:1:1:63:1:4:0:11:10

to play 63

1.	XG Roller++	23/17 5/2	eq: +0.658
	Player: Opponent: 58.02% (G:22.91% B:1.07%) 41.98% (G:13.73% B:0.75%)
2.	XG Roller++	10/4 5/2	eq: +0.400 (-0.258)
	Player: Opponent: 49.63% (G:19.35% B:0.74%) 50.37% (G:14.56% B:0.70%)
3.	2-ply	8/5 8/2	eq: +0.349 (-0.309)
	Player: Opponent: 48.47% (G:18.96% B:0.83%) 51.53% (G:15.08% B:0.58%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I was terrified of getting hit here so played 10/4, 6/3. My thinking was why give extra shots when I have a strong cube next time on plenty of rolls?

Apparently the motto was "put em where you want em" when opponent is on bar versus 4 point board. 23/18, 6/3 giving 16, 36, and 33 as extras is the computer move.

Apparently the position after my play and dance isn't as strong as it looks. Is Redouble / Easy take. Perhaps this is influencing the play. Take a calculated risk now to get a better cube next time.

***********

I lead 4 away 7 away and am on roll. Cube action?

is David Todd score: 4 pip: 142

11 point match

pip: 127 score: 7 is Bill

XGID=-b-B-bC-CB-B-C---a-cbcb---:0:0:1:00:7:4:0:11:10

on roll, cube action?

Analyzed in XG Roller++	No double	Double/Take
Player Winning Chances:	66.37% (G:9.25% B:0.35%)	66.55% (G:9.24% B:0.33%)
Opponent Winning Chances:	33.63% (G:7.12% B:0.12%)	33.45% (G:6.66% B:0.09%)
Cubeless Equities	+0.376	+0.932
Cubeful Equities
No double:	+0.605
Double/Take:	+0.437 (-0.168)
Double/Pass:	+1.000 (+0.395)
Best Cube action: No double / Take
Percentage of wrong pass needed to make the double decision right: 23.0%

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I went for it and doubled. I was rewarded with a drop but apparently this is a no double. Also no double for money.

Weird. i don't get it. It looks like he has no timing so my position is just very strong. OK so DT can't play a backgame but this isn't the only game plan. He has a death board waiting for me so I will ahve to tiptoe this game home. Considering my lame board and 10 outside men this certainly won't be easy. DT will probably be able to scoot his back man up and hold the 5 point. The race is not hopeless - he can probably win once in a while in the race. Most likely he holds the 5 point and gives me problems bearing in. Even if he has to abandong the 5 point he will have play from the ace point anchor. I still have plenty of gaps in my home board to fill in. And any hit will probably be a game winner now and well into the future.


***************

I lead 3 away 7 away and have 52 to play owning the cube.

is David Todd score: 4 pip: 137

11 point match

pip: 124 score: 8 is Bill

XGID=aaBB--DAC--Ab--b---cbbb-B-:1:1:1:52:8:4:0:11:10

to play 52

1.	XG Roller++	8/1*	eq: -0.391
	Player: Opponent: 39.56% (G:21.45% B:0.30%) 60.44% (G:23.91% B:2.76%)
2.	XG Roller++	11/9 7/2	eq: -0.612 (-0.221)
	Player: Opponent: 29.27% (G:10.96% B:0.37%) 70.73% (G:22.42% B:1.27%)
3.	XG Roller+	11/9 6/1*	eq: -0.423 (-0.032)
	Player: Opponent: 39.53% (G:21.41% B:0.31%) 60.47% (G:24.17% B:2.64%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I was possesed by fear. 11/9, 7/2 with blot control (and gammon-phobia) on my mind.

8/1* was the way to go. "Two on the bar is better by far." And my game plan, as unlikely as it is, is to attack and try to escape later.

My wimpy play simply hopes to not get gammoned. This is another case of growing a pair and playing to win rather than playing not to lose.